Use this URL to cite or link to this record in EThOS:
Title: Resilient routing in the internet
Author: Sae Lor, S.
ISNI:       0000 0004 2728 8630
Awarding Body: University College London (University of London)
Current Institution: University College London (University of London)
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
Access from Institution:
Although it is widely known that the Internet is not prone to random failures, unplanned failures due to attacks can be very damaging. This prevents many organisations from deploying beneficial operations through the Internet. In general, the data is delivered from a source to a destination via a series of routers (i.e routing path). These routers employ routing protocols to compute best paths based on routing information they possess. However, when a failure occurs, the routers must re-construct their routing tables, which may take several seconds to complete. Evidently, most losses occur during this period. IP Fast Re-Route (IPFRR), Multi-Topology (MT) routing, and overlays are examples of solutions proposed to handle network failures. These techniques alleviate the packet losses to different extents, yet none have provided optimal solutions. This thesis focuses on identifying the fundamental routing problem due to convergence process. It describes the mechanisms of each existing technique as well as its pros and cons. Furthermore, it presents new techniques for fast re-routing as follows. Enhanced Loop-Free Alternates (E-LFAs) increase the repair coverage of the existing techniques, Loop-Free Alternates (LFAs). In addition, two techniques namely, Full Fast Failure Recovery (F3R) and fast re-route using Alternate Next Hop Counters (ANHC), offer full protection against any single link failures. Nevertheless, the former technique requires significantly higher computational overheads and incurs longer backup routes. Both techniques are proved to be complete and correct while ANHC neither requires any major modifications to the traditional routing paradigm nor incurs significant overheads. Furthermore, in the presence of failures, ANHC does not jeopardise other operable parts of the network. As emerging applications require higher reliability, multiple failures scenarios cannot be ignored. Most existing fast re-route techniques are able to handle only single or dual failures cases. This thesis provides an insight on a novel approach known as Packet Re-cycling (PR), which is capable of handling any number of failures in an oriented network. That is, packets can be forwarded successfully as long as a path between a source and a destination is available. Since the Internet-based services and applications continue to advance, improving the network resilience will be a challenging research topic for the decades to come.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available